Can we shift belief in the 'Law of Small Numbers'?

'Sample size neglect' is a tendency to underestimate how the variability of mean estimates changes with sample size. We studied 100 participants, from science or social science backgrounds, to test whether a training task showing different-sized samples of data points (the 'beeswarm' task) can help overcome this bias. Ability to judge if two samples came from the same population improved with training, and 38% of participants reported that they had learned to wait for larger samples before making a response. Before and after training, participants completed a 12-item estimation quiz, including items testing sample size neglect (S-items). Bonus payments were given for correct responses. The quiz confirmed sample size neglect: 20% of participants scored zero on S-items, and only two participants achieved more than 4/6 items correct. Performance on the quiz did not improve after training, regardless of how much learning had occurred on the beeswarm task. Error patterns on the quiz were generally consistent with expectation, though there were some intriguing exceptions that could not readily be explained by sample size neglect. We suggest that training with simulated data might need to be accompanied by explicit instruction to be effective in counteracting sample size neglect more generally.

An updated investigation of the multidimensional structure of language lateralization in left- and right-handed adults: a test-retest functional transcranial Doppler sonography study with six language tasks.

A previous study we reported in this journal suggested that left and right-handers may differ in their patterns of lateralization for different language tasks (Woodhead et al. 2019 R. Soc. Open Sci. 6, 181801. (doi:10.1098/rsos.181801)). However, it had too few left-handers (N = 7) to reach firm conclusions. For this update paper, further participants were added to the sample to create separate groups of left- (N = 31) and right-handers (N = 43). Two hypotheses were tested: (1) that lateralization would be weaker at the group level in left-than right-handers; and (2) that left-handers would show weaker covariance in lateralization between tasks, supporting a two-factor model. All participants performed the same protocol as in our previous paper: lateralization was measured using functional transcranial Doppler sonography during six different language tasks, on two separate testing sessions. The results supported hypothesis 1, with significant differences in laterality between groups for four out of six tasks. For hypothesis 2, structural equation modelling showed that there was stronger evidence for a two-factor model in left than right-handers; furthermore, examination of the factor loadings suggested that the pattern of laterality across tasks may also differ between handedness groups. These results expand on what is known about the differences in laterality between left- and right-handers.

Investigating the effects of handedness on the consistency of lateralization for speech production and semantic processing tasks using functional transcranial Doppler sonography.

The left hemisphere is dominant for language in most people, but lateralization strength varies between different tasks and individuals. A large body of literature has shown that handedness is associated with lateralization: left handers have weaker language lateralization on average, and a greater incidence of atypical (right hemisphere) lateralization; but typically, these studies have relied on a single measure of language lateralization. Here we consider the relationships between lateralization for two different language tasks. We investigated the influence of handedness on lateralization using functional transcranial Doppler sonography (fTCD), using an existing dataset (N = 151 adults, 21 left handed). We compared a speech production task (word generation) and a semantic association task. We demonstrated stronger left-lateralization for word generation than semantic association; and a moderate correlation between laterality indices for the two tasks (r = 0.59). Laterality indices were stronger for right than left handers, and left handers were more likely than right handers to have atypical (right hemisphere) lateralization or inconsistent lateralization between the two tasks. These results add to our knowledge of individual differences in lateralization and support the view that language lateralization is multifactorial rather than unitary.

Psychoanalysis in the treatment of autism: why is France a cultural outlier?

In most countries, social or behavioural interventions are recommended for autism. However, in France, psychoanalysis is still used, despite objections by patients, families and mental health experts. Supporters of psychoanalysis maintain that the choice of therapeutic approach is a matter of cultural preference, and that objections to psychoanalysis arise from misunderstandings. We argue that more deep-rooted problems are the lack of an evidence base for psychoanalysis and its focus on sexual relationships between children and adults, which demonises mothers and can put children at risk of abuse. Furthermore, psychoanalysis in France is protected from criticism by powerful educational and political networks.

Judging meaning: A domain-level difference between autistic and non-autistic adults.

We tested whether autistic adults would show selective difficulties across several tests of inferencing and social understanding in the context of average-range core language ability. One-hundred and ninety-one participants completed an online battery, and data were analysed using confirmatory factor analysis. Results showed that vocabulary knowledge was separate from other measures, which collectively formed a 'receptive communication' factor. Autistic people underperformed on the 'receptive communication' factor but showed more advanced vocabulary knowledge than non-autistic participants. Lower performance among autistic adults on the test battery predicted face-to-face communication difficulties measured by self-report and researcher ratings, with moderate effect sizes. Follow-up analysis indicated three further findings. We hypothesized that differences would arise from an isolated 'theory of mind' difficulty in autistic people, but instead the data suggested more general information-processing differences when making judgements about communicative stimuli. Second, substantial group differences on a test of implied meaning were only partly explained at the factor level, suggesting that multiple cognitive influences underpinned these differences. Finally, autistic women tended to perform better than autistic men. Our results support the idea of a subtle domain-level difference in pragmatics in autistic people, while questioning the basis of this difference and highlighting substantial variability in skills across groups.

Testing the unitary theory of language lateralization using functional transcranial Doppler sonography in adults.

Hemispheric dominance for language can vary from task to task, but it is unclear if this reflects error of measurement or independent lateralization of different language systems. We used functional transcranial Doppler sonography to assess language lateralization within the middle cerebral artery territory in 37 adults (seven left-handers) on six tasks, each given on two occasions. Tasks taxed different aspects of language function. A pre-registered structural equation analysis was used to compare models of means and covariances. For most people, a single lateralized factor explained most of the covariance between tasks. A minority, however, showed dissociation of asymmetry, giving a second factor. This was mostly derived from a receptive task, which was highly reliable but not lateralized. The results suggest that variation in the strength of language lateralization reflects true individual differences and not just error of measurement. The inclusion of several tasks in a laterality battery makes it easier to detect cases of atypical asymmetry.

CATALISE: A Multinational and Multidisciplinary Delphi Consensus Study. Identifying Language Impairments in Children.

Delayed or impaired language development is a common developmental concern, yet there is little agreement about the criteria used to identify and classify language impairments in children. Children's language difficulties are at the interface between education, medicine and the allied professions, who may all adopt different approaches to conceptualising them. Our goal in this study was to use an online Delphi technique to see whether it was possible to achieve consensus among professionals on appropriate criteria for identifying children who might benefit from specialist services. We recruited a panel of 59 experts representing ten disciplines (including education, psychology, speech-language therapy/pathology, paediatrics and child psychiatry) from English-speaking countries (Australia, Canada, Ireland, New Zealand, United Kingdom and USA). The starting point for round 1 was a set of 46 statements based on articles and commentaries in a special issue of a journal focusing on this topic. Panel members rated each statement for both relevance and validity on a seven-point scale, and added free text comments. These responses were synthesised by the first two authors, who then removed, combined or modified items with a view to improving consensus. The resulting set of statements was returned to the panel for a second evaluation (round 2). Consensus (percentage reporting 'agree' or 'strongly agree') was at least 80 percent for 24 of 27 round 2 statements, though many respondents qualified their response with written comments. These were again synthesised by the first two authors. The resulting consensus statement is reported here, with additional summary of relevant evidence, and a concluding commentary on residual disagreements and gaps in the evidence base.

Open research practices: unintended consequences and suggestions for averting them. (Commentary on the Peer Reviewers' Openness Initiative).

The Peer Reviewers' Openness Initiative (PROI) is a move to enlist reviewers in the promotion of data-sharing. In this commentary, I discuss objections that can be raised, first to the specific proposals in the PROI, and second to data-sharing in general. I argue that although many objections have strong counter-arguments, others merit more serious consideration. Regarding the PROI, I suggest that it could backfire if editors and authors feel coerced into data-sharing and so may not be the most pragmatic way of encouraging greater openness. More generally, while promoting data-sharing, we need to be sensitive to cases where sharing of data from human participants could create ethical problems. Furthermore, those interested in promoting reproducible science need to defend against an increased risk of data-dredging when large, multivariable datasets are shared. I end with some suggestions to avoid these unintended consequences.

The interface between genetics and psychology: lessons from developmental dyslexia.

Developmental dyslexia runs in families, and twin studies have confirmed that there is a substantial genetic contribution to poor reading. The way in which discoveries in molecular genetics are reported can be misleading, encouraging us to think that there are specific genes that might be used to screen for disorder. However, dyslexia is not a classic Mendelian disorder that is caused by a mutation in a single gene. Rather, like many other common disorders, it appears to involve combined effects of many genes and environmental factors, each of which has a small influence, possibly supplemented by rare variants that have larger effects but apply to only a minority of cases. Furthermore, to see clearer relationships between genotype and phenotype, we may need to move beyond the clinical category of dyslexia to look at underlying cognitive deficits that may be implicated in other neurodevelopmental disorders.

Ten questions about terminology for children with unexplained language problems.

BACKGROUND: In domains other than language, there is fairly consistent diagnostic terminology to refer to children's developmental difficulties. For instance, the terms 'dyslexia', 'attention deficit hyperactivity disorder' and 'autistic spectrum disorder' are used for difficulties with reading, attention or social cognition, respectively. There is no agreed label, however, for children with unexplained language problems. AIMS: To consider whether we need labels for unexplained language problems in children, and if so, what terminology is appropriate. MAIN CONTRIBUTION: There are both advantages and disadvantages to labels, but they are important to ensure children receive services, and to increase our knowledge of the nature and causes of such problems. A survey of labels in current use found 132 different terms, 33 of which had 600 or more returns on Google Scholar between 1994 and 2013. Many of these labels were too general to be useful. Of the remainder, the term 'specific language impairment' was the most commonly used. CONCLUSIONS: The current mayhem in diagnostic labels is unsustainable; it causes confusion and impedes research progress and access to appropriate services. We need to achieve consensus on diagnostic criteria and terminology. The DSM-5 term 'language disorder' is problematic because it identifies too wide a range of conditions on an internet search. One solution is to retain specific language impairment, with the understanding that 'specific' means idiopathic (i.e., of unknown origin) rather than implying there are no other problems beyond language. Other options are the terms 'primary language impairment', 'developmental language disorder' or 'language learning impairment'.

Research Review: Emanuel Miller Memorial Lecture 2012 - neuroscientific studies of intervention for language impairment in children: interpretive and methodological problems.

BACKGROUND: Our ability to look at structure and function of a living brain has increased exponentially since the early 1970s. Many studies of developmental disorders now routinely include a brain imaging or electrophysiological component. Amid current enthusiasm for applications of neuroscience to educational interventions, we need to pause to consider what neuroimaging data can tell us. Images of brain activity are seductive, and have been used to give credibility to commercial interventions, yet we have only a limited idea of what the brain bases of language disorders are, let alone how to alter them. SCOPE AND FINDINGS: A review of six studies of neuroimaging correlates of language intervention found recurring methodological problems: lack of an adequate control group, inadequate power, incomplete reporting of data, no correction for multiple comparisons, data dredging and failure to analyse treatment effects appropriately. In addition, there is a tendency to regard neuroimaging data as more meaningful than behavioural data, even though it is behaviour that interventions aim to alter. CONCLUSION: In our current state of knowledge, it would be better to spend research funds doing well-designed trials of behavioural treatment to establish which methods are effective, rather than rushing headlong into functional imaging studies of unproven treatments.

Auditory deficit as a consequence rather than endophenotype of specific language impairment: electrophysiological evidence.

BACKGROUND: Are developmental language disorders caused by poor auditory discrimination? This is a popular theory, but behavioural evidence has been inconclusive. Here we studied children with specific language impairment, measuring the brain's electrophysiological response to sounds in a passive paradigm. We focused on the T-complex, an event-related peak that has different origins and developmental course from the well-known vertex response. METHODS: We analysed auditory event-related potentials to tones and syllables from 16 children and 16 adolescents with specific language impairment who were compared with 32 typically-developing controls, matched for gender, IQ and age. RESULTS: We replicated prior findings of significant reduction in Ta amplitude for both children and adolescents with specific language impairment, which was particularly marked for syllables. The topography of the T-complex to syllables indicated a less focal response in those with language impairments. To distinguish causal models, we considered correlations between size of the Ta response and measures of language and literacy in parents as well as children. The best-fitting model was one in which auditory deficit was a consequence rather than a cause of difficulties in phonological processing. CONCLUSIONS: The T-complex to syllables has abnormal size and topography in children with specific language impairment, but this is more likely to be a consequence rather than a cause of difficulties in phonological processing.

CNTNAP2 variants affect early language development in the general population.

Early language development is known to be under genetic influence, but the genes affecting normal variation in the general population remain largely elusive. Recent studies of disorder reported that variants of the CNTNAP2 gene are associated both with language deficits in specific language impairment (SLI) and with language delays in autism. We tested the hypothesis that these CNTNAP2 variants affect communicative behavior, measured at 2 years of age in a large epidemiological sample, the Western Australian Pregnancy Cohort (Raine) Study. Singlepoint analyses of 1149 children (606 males and 543 females) revealed patterns of association which were strikingly reminiscent of those observed in previous investigations of impaired language, centered on the same genetic markers and with a consistent direction of effect (rs2710102, P = 0.0239; rs759178, P = 0.0248). On the basis of these findings, we performed analyses of four-marker haplotypes of rs2710102-rs759178-rs17236239-rs2538976 and identified significant association (haplotype TTAA, P = 0.049; haplotype CGAG, [corrected] P = .0014). Our study suggests that common variants in the exon 13-15 region of CNTNAP2 influence early language acquisition, as assessed at age 2, in the general population. We propose that these CNTNAP2 variants increase susceptibility to SLI or autism when they occur together with other risk factors.

Overlaps between autism and language impairment: phenomimicry or shared etiology?

Traditionally, autistic spectrum disorder (ASD) and specific language impairment (SLI) are regarded as distinct conditions with separate etiologies. Yet these disorders co-occur at above chance levels, suggesting shared etiology. Simulations, however, show that additive pleiotropic genes cannot account for observed rates of language impairment in relatives, which are higher for probands with SLI than for those with ASD + language impairment. An alternative account is in terms of 'phenomimicry', i.e., language impairment in comorbid cases may be a consequence of ASD risk factors, and different from that seen in SLI. However, this cannot explain why molecular genetic studies have found a common risk genotype for ASD and SLI. This paper explores whether nonadditive genetic influences could account for both family and molecular findings. A modified simulation involving G x G interactions obtained levels of comorbidity and rates of impairment in relatives more consistent with observed values. The simulations further suggest that the shape of distributions of phenotypic trait scores for different genotypes may provide evidence of whether a gene is involved in epistasis.

Measurement of mismatch negativity in individuals: a study using single-trial analysis.

Mismatch negativity (MMN) is measured by subtracting the averaged response to a set of standard stimuli from the averaged response to rarer deviant stimuli, and taking the amplitude of this difference wave in a given time window. This method is problematic when used to evaluate individuals, because there is no estimate of variance. We describe a new approach, in which independent components with high trial-by-trial variance are first removed. Next, each deviant response has the preceding standard response subtracted, giving a set of single trial difference waves. We illustrate this approach in analysis of MMN to brief tones in 17 adults. The best criterion for MMN combined t-test with an index of inter-trial coherence, giving significant MMN in 14 (82%) of individuals. Single-trial methods can indicate which people show MMN. However, in some clinically normal individuals there was no MMN, despite good behavioral discrimination of stimuli.

Genes, cognition, and communication: insights from neurodevelopmental disorders.

Twin and family studies have demonstrated that most cognitive traits are moderately to highly heritable. Neurodevelopmental disorders such as dyslexia, autism, and specific language impairment (SLI) also show strong genetic influence. Nevertheless, it has proved difficult for researchers to identify genes that would explain substantial amounts of variance in cognitive traits or disorders. Although this observation may seem paradoxical, it fits with a multifactorial model of how complex human traits are influenced by numerous genes that interact with one another, and with the environment, to produce a specific phenotype. Such a model can also explain why genetic influences on cognition have not vanished in the course of human evolution. Recent linkage and association studies of SLI and dyslexia are reviewed to illustrate these points. The role of nonheritable genetic mutations (sporadic copy number variants) in causing autism is also discussed. Finally, research on phenotypic correlates of allelic variation in the genes ASPM and microcephalin is considered; initial interest in these as genes for brain size or intelligence has been dampened by a failure to find phenotypic differences in people with different versions of these genes. There is a current vogue for investigators to include measures of allelic variants in studies of cognition and cognitive disorders. It is important to be aware that the effect sizes associated with these variants are typically small and hard to detect without extremely large sample sizes.

Weak hand preference in children with down syndrome is associated with language deficits.

This study explores associations between language ability and hand preference in children with Down syndrome. Compared to typically developing children of the same age, children with Down syndrome showed weaker hand preference, were less consistent in the hand they used and also less willing to reach to extreme positions in contralateral space. Within the group of children with Down syndrome, those who showed a stronger or more consistent hand preference had better language and memory skills. This association could not be explained by differences in non-verbal cognitive ability or hearing loss. These findings are discussed within the theory of neurolinguistic development proposed by Locke [Locke (1997). Brain & Language, 58, 265-326].

Heritability of specific language impairment depends on diagnostic criteria.

Heritability estimates for specific language impairment (SLI) have been inconsistent. Four twin studies reported heritability of 0.5 or more, but a recent report from the Twins Early Development Study found negligible genetic influence in 4-year-olds. We considered whether the method of ascertainment influenced results and found substantially higher heritability if SLI was defined in terms of referral to speech and language pathology services than if defined by language test scores. Further analysis showed that presence of speech difficulties played a major role in determining whether a child had contact with services. Childhood language disorders that are identified by population screening are likely to have a different phenotype and different etiology from clinically referred cases. Genetic studies are more likely to find high heritability if they focus on cases who have speech difficulties and who have been referred for intervention.

Using mismatch negativity to study central auditory processing in developmental language and literacy impairments: where are we, and where should we be going?

A popular theoretical account of developmental language and literacy disorders implicates poor auditory temporal processing in their etiology, but evidence from studies using behavioral measures has yielded inconsistent results. The mismatch negativity (MMN) component of the auditory event-related potential has been recommended as an alternative, relatively objective, measure of the brain's ability to discriminate sounds that is suitable for children with limited attention or motivation. A literature search revealed 26 studies of the MMN in individuals with dyslexia or specific language impairment and 4 studies of infants or children at familial risk of these disorders. Findings were highly inconsistent. Overall, attenuation of the MMN and atypical lateralization in the clinical group were most likely to be found in studies using rapidly presented stimuli, including nonverbal sounds. The MMN literature offers tentative support for the hypothesis that auditory temporal processing is impaired in language and literacy disorders, but the field is plagued by methodological inconsistencies, low reliability of measures, and low statistical power. The article concludes with recommendations for improving this state of affairs.

Heritable risk factors associated with language impairments.

There is a strong genetic contribution to children's language and literacy impairments. The aim of this study was to determine which aspects of the phenotype are familial by comparing 34 parents of probands with language/literacy impairments and 33 parents of typically developing probands. The parents responded to questionnaires regarding previous history for language/reading impairment and participated in psychometric testing. The psychometric test battery consisted of tests assessing non-verbal IQ, short-term memory, articulation, receptive grammar, reading abilities and spelling. Self-report measures demonstrated a higher prevalence of language and literacy impairments in parents of affected probands (32%) compared with parents of unaffected probands (6%). The two groups of parents differed significantly in their performance on the non-word repetition, oromotor and digit span tasks. Non-word repetition gave the best discrimination between the parent groups even when the data from the parents who actually were impaired as ascertained by direct testing or self-report were removed from the analyses. This suggests that non-word repetition serves as a marker of a family risk for language impairment. The paper concludes with a discussion of issues associated with ascertainment of specific language impairment (SLI).